Chronic kidney disease (CKD) affects approximately 10% of the general population and is especially common among US veterans, affecting up to 36% of all VA users during FY2006-2014. All-cause and cardiovascular (CV) morbidity and mortality are extremely high in veterans with CKD and ESRD, and African American veterans have substantially higher rates of CKD and CV disease than whites. Traditional risk factors of cardiovascular disease such as hypercholesterolemia, hypertension and obesity often show seemingly anomalous, inverse associations with adverse outcomes among individuals with CKD and ESRD. To date there are few therapeutic interventions proven to prevent the development of CKD, delay its progression, or decrease morbidity/mortality in patients with CKD. There is also a lack of race-specific application of clinical interventions, in spite of evidence suggesting that African-Americans may respond differently to various therapies. The lack of clinical end point driven randomized controlled trials in this population necessitates robust epidemiologic studies to provide preliminary results in support of future clinical trials, and to offer large- scale, widely applicable evidence to inform clinical practice in cases when clinical trials are not feasible. Our proposal will expand on our previous work using data obtained from the national VA research database, which offers uniquely detailed sociodemographic and clinical information on millions of individuals across all parts of the US. The Specific Aims of our proposal are: Aim 1: Examine the effect of interventions on clinical outcomes in veterans with non-dialysis dependent CKD (NDD-CKD) overall and by race-ethnicity. Aim 2: Examine the effect of interventions on incident CKD and on all-cause and cause-specific mortality, CHD, incident stroke and incident CHF in veterans with normal baseline estimated GFR, overall and by race- ethnicity. Aim 3: Employ powerful graph theoretical algorithms and scalable supercomputer implementations to test and extend the validity of the findings from Aims 1 and 2, and to help elucidate hidden factors and highlight previously unknown relationships between condition, risk, treatment and outcome. Next steps: This four-year project will generate a wealth of information to examine the above outcomes of patients with all levels of kidney function, and could have significant implications for the care of US veterans and also for patients with kidney diseases in general. Clinicians and guideline committees could use information generated from our research to determine the most likely benefits for interventions without clinical trial evidence, and this could help prioritization of healthcare expenditures and drug formularies. Clinical trialists could also use information generated from our research to determine likely effect sizes for various interventions and various subgroups that may be most prone to benefit from interventions. Furthermore,